Cowan Kyra J, Erickson Rich, Sue Brian, Delarosa Reginald, Gunter Bert, Coleman Daniel A, Gilbert Houston, Song An, Fischer Saloumeh Kadkhodayan
BioAnalytical Sciences, Genentech, 1 DNA Way, South San Francisco, CA 94080, USA.
Bioanalysis. 2012 Sep;4(17):2127-39. doi: 10.4155/bio.12.199.
Design of experiments (DOE) is a systematic approach to assess the effects of many factors on a response of an assay. This paper provides a case study whereby DOE was successfully utilized to evaluate robustness parameters for a ligand-binding assay (LBA).
A 24-run Plackett-Burman design was developed to investigate factors that may have caused a lack of robustness in this particular LBA. We modeled five main effects and their ten two-way interactions, using the standard curve signal as the response.
By utilizing DOE, we were able to quickly identify the factors that affected our assay's performance. The lack of robustness was attributed to the handling of the coat reagent. Factors that had an adverse effect on the coat material were vortexing and freeze-thaw cycles.
We recommend that a robustness DOE be conducted prior to the validation of an assay for early identification of critical factors that may impact assay performance.
